The Ranking System
The Ranking System
The ELO Model
Elo modelling is a commonly-used approach towards creating ranking systems in sport originally derived for chess. It is used across multiple sports at international level including golf, tennis and snow-boarding.
The principle is that points are awarded for each event based on your previous ranking and the difference between you and the teams you beat.
• If a lower ranked team (A) beats a higher ranked team (B), team A is rewarded a lot because they were expected to lose, but won. Likewise, team B loses points because they were expected to win, but lost.
• If a lower ranked team (A) loses to a higher ranked team (B), team A loses less points because they were expected to lose and lost. Team B should not be rewarded much because they were expected to win and won.
This prevents higher ranking teams continually racing lower ranking teams just to gain ‘easy’ points.
The main purpose of this is to rank the teams from highest to lowest performing and not to award points, however to get a ranking points are calculated in accordance with the following equations.
Most people can stop reading now!
For the mathematically minded here is the detail
RW = Elo rating of wining boat,
RL = Elo rating of losing boat,
EW = Expected points of boat 1 = 1 1+10(𝑅𝐿−𝑅𝑊)/400 EL = Expected points of losing = 1 1+10(𝑅𝑊−𝑅𝐿)/400
and K (The k factor),also referred to as the rate of change, a parameter defined for each event.
The winning boat’s new rating calculated as: RW + K*(1 – EW)
The losing boat’s new rating is calculated as: RL + K*(0 – EL)
For multiple boats in an event we are implementing a multi-elo model. We are using initial parameters of k = 32 and d = 400 based on initial running of the 2019 models. If anyone has any expertise in this area we would love to hear from you